This project involves the creation, development, and testing of image processing techniques designed to improve diagnostic performance. Of continued interest is a computer implemented transformation which generates an isometric three-dimensional "pseudo-solid" from any two-dimensional array of grey levels. This redundant coding scheme permits an image to be viewed as a projection of an opaque solid whose shape is determined by the original distribution of grey levels. When applied to dental radiographs induced interproximal lesions were enhanced in a way which is relatively independent of the information capacity of the image. Projection angle of the simulated source of illumination and viewing geometry were found to be task-dependent and sufficiently independent to permit rational bases for optimization. A new enhancement scheme derived from second and third order estimates of image redundancy is being explored as a possible basis for extending the principle of histogram equalization. This scheme is a logical extension of information theory which may aid in image abstraction as well as analysis.